Abstract

Intense domestic physical activity (IDPA) is promoted by preventive health campaigns, but this recommendation is not supported by evidence. The authors used data from the 1995, 1998, and 2003 Scottish Health Survey samples and the associated mortality and hospital episode records to determine the independent effects of IDPA on cardiovascular disease (CVD) events and all-cause mortality. The sample comprised 13,726 (6,102 men) CVD-free respondents (≥35 years). Multivariable survival analysis assessed the relation between IDPA and the risk for CVD (fatal/nonfatal combined) or all-cause mortality. During 8.4 (standard deviation, 3.4) years of follow-up, there were 1,103 deaths (573 among men) and 890 CVD events (521 among men). Participation in IDPA was associated with lower all-cause mortality (men: relative risk = 0.68, 95% confidence interval: 0.50, 0.91; women: relative risk = 0.70, 95% confidence interval: 0.52, 0.93). In both sexes, IDPA was unrelated to the risk for CVD. Total physical activity (including IDPA) was unrelated to fatal/nonfatal CVD, but when domestic activity was excluded from the calculations there was an association (men: relative risk = 0.76, 95% confidence interval: 0.58, 0.98; women: relative risk = 0.68, 95% confidence interval: 0.50, 0.93). These results indicate that IDPA may not offer protection against CVD, but it may protect against all-cause mortality. CVD preventive efforts may need to focus on moderate-to-vigorous-intensity physical activities other than those performed in and around the household.

Scotland has by far the highest rate of heart disease-related deaths in the United Kingdom and one of the highest rates of coronary heart disease events in the world (1). There is indisputable evidence that regular physical activity protects against chronic disease and premature death and, in particular, against cardiovascular disease (CVD) (2). Current physical activity recommendations advocate the accumulation of at least 30 minutes of moderate-to-vigorous physical activity on at least 5 days a week (3–5). One of the key messages is that activities performed during daily living (e.g., walking or cycling, stair climbing, gardening, intense housework) may be sufficient to prevent chronic disease and premature death. Although the evidence on the health benefits of walking, cycling, or stair climbing is compelling (6–13), the degree of protection that intense domestic physical activity (IDPA) confers is much less clear.

In agreement with previous cross-sectional observations (10, 14), our findings recently showed (9) that IDPA does not relate independently to known CVD risk factors. Few prospective studies (6, 7, 15–17) have considered domestic activity in relation to mortality or CVD risk, but the evidence as a whole is inconclusive because domestic activity is analyzed as part of overall lifestyle activity, making it difficult to disentangle its independent effects. Despite this paucity of evidence, United Kingdom (3) and US (4) public health recommendations made repeated references to participation in IDPA as a means of acquiring the required amounts of physical activity to prevent CVD. This is in line with other primary care (18) and public health (19) recommendations that encouraged engagement in IDPA to achieve health benefits. IDPA is highly prevalent in Britain (9, 14), and its exclusion reduces the percentage of adults meeting the current physical activity recommendations by approximately 30% (9). In addition to a lack of robust epidemiologic evidence supporting a cardioprotective effect of IDPA, there is physiologic support that cardioprotective physical activity should be characterized by the use of large muscle groups utilized in a rhythmic or dynamic nature (20). IDPA, by contrast, usually utilizes smaller upper body muscles and may be more intermittent and less rhythmic.

It is important that this literature gap be addressed, because participation in IDPA may prevent participation in other, potentially more health-enhancing physical activity as the result of time constraints and individuals’ believing the health-promoting messages that this is sufficient activity for health enhancement.

The aim of this study was therefore to examine the independent effects of IDPA on all-cause mortality CVD events.

MATERIALS AND METHODS

Study population

The Scottish Health Survey consists of cross-sectional general population-based surveys among individuals living in households in Scotland. The 1995, 1998, and 2003 Scottish Health Survey samples included in this analysis were selected by using a multistage stratified probability sampling design to give a representative sample of the target population. Further details on sample design and selection can be found elsewhere (21). Ethics approval had been granted by the Local Research Ethics Councils. Scottish Health Survey data were collected approximately evenly throughout the year. Adult response rates (percentage of those eligible) were 81% in 1995, 77% in 1998, and 83% in 2003. Data were collected by survey interviewers during household visits. This study considered 16,144 respondents (7,211 male), representing approximately 92% of the respondents in each survey year who consented to their records being linked to National Health Service administrative data. Included age ranges at baseline were 35–64 years from the 1995 survey, 35–74 years in 1998, and over 35 years in 2003. The surveys were linked to the patient-based database of hospital episodes and deaths up to December 2007. Hospital episodes data were available from 1980 onward. We considered hospital episodes with a principal diagnosis of CVD. Diagnoses for primary cause of death were recorded according to the International Classification of Diseases, Ninth Revision and Tenth Revision, with CVD codes 390–459 from the Ninth Revision and I01–I99 from the Tenth Revision. All analyses excluded respondents who had CVD hospital episodes between 1981 and the baseline measurement to minimize the possibility that changes in physical activity following the episode and diagnosis would obscure the true association between physical activity and risk for CVD events or death. The all-cause mortality analysis excluded respondents with a cancer registration prior to the measurement for the same reason. Of the 16,144 participants who completed the physical activity interview, 1,735 respondents had missing covariables, 198 had missing physical activity values, and 726 (485 men, 241 women) had CVD at baseline and were excluded from analyses. The resulting 13,726 individuals comprised the core data set (6,102 men, 7,624 women). From this data set, for the analyses with all-cause mortality, we excluded 456 respondents (181 men, 275 women) because they had been diagnosed with cancer prior to baseline assessment. This resulted in 13,221 individuals (5,896 men, 7,325 women) being included in the all-cause mortality analyses. For the analyses with CVD as the outcome, we excluded (from the core data set) 132 respondents with a fatal or nonfatal CVD event during the first year of follow-up (68 men, 64 women), resulting in 13,594 individuals (6,033 men, 7,561 women) being included in the CVD analyses.

For the 8.4 (standard deviation, 3.4) years of average follow-up, there were 114,838 person-years at risk. In the core sample, there were 573 any-cause deaths among men (184 due to CVD causes), 530 any-cause deaths among women (143 due to CVD), and a total of 890 CVD fatal/nonfatal events (521 in men, 369 in women). After all the exclusions described above, 944 (499 in men, 445 in women) deaths occurred in the sample entered in the all-cause mortality analysis, and 794 CVD events (468 in men, 326 in women) occurred in the CVD analysis sample.

Physical activity

In 1998 and 2003, the physical activity questions included frequency (number of days in the last 4 weeks) and duration (minutes per day) of participation in heavy housework (e.g., scrubbing floors, cleaning windows), heavy “do-it-yourself” activities/gardening (e.g., sweeping leaves, digging, building work), walking for any purpose, and any leisure-time sports and exercises (e.g., cycling, swimming, aerobics, calisthenics, gym, dancing, football or rugby, racket sports). The 1995 physical activity questions inquired about the duration and frequency of physical activity during an average week. The 1995 questions covered the same physical activity domains as in 1998 and 2003, but duration and frequency were reported as close-ended 5-point categorical variables ranging from no participation to 6–7 times a week (frequency) and from no time to 2 hours or longer for duration. To make the data from 1998–2003 and 1995 compatible, we converted the 1995 categorical frequency and duration responses to continuous variables by taking the midpoint of each category. For example, “2–3 times a week” was set to 2.5 times, and “20 minutes, less than 30 minutes” was set to 25 minutes. We also divided the 1998–2003 frequency and duration responses by 4 to elicit average weekly sessions. The 1995 questions inquired about the usual breathing rate of walking (normal, faster than normal, gasping for breath), while walking intensity in 1998–2003 was measured by a question on walking pace (slow, steady average, fairly brisk, fast). Those 2 questions elicited very different prevalences in each category and made it impossible for us to standardize across both assessments (i.e., 11% included in the highest 2 categories—breathing faster than normal or gasping for breath—of the first assessment and 31% included in the highest 2 categories of the second measurement); for this reason, we included all walking in our analyses, regardless of perceived intensity.

The criterion validity of the physical activity questions used in 1998 and 2003 is supported by a recent study on 106 general population English adults (45 men), where the output of accelerometers (worn for 2 nonconsecutive weeks over a 1-month period) was compared against a slightly modified version of the above questions (22). The questionnaire appeared to be a valid measure of the time spent in moderate-to-vigorous physical activity, with intraclass correlation coefficients of 0.47 in men (P = 0.03) and 0.43 in women (P = 0.02). In terms of test-retest reliability, the coefficients of time spent in moderate-to-vigorous physical activity were 0.89 in men (P < 0.001) and 0.76 in women (P < 0.001). Additionally, these physical activity measures have demonstrated excellent convergent validity in grading a plethora of biochemical and physiologic CVD risk factors by nondomestic physical activity types, such as walking and sports (9).

Anthropometry and other factors

Height and weight were measured by trained interviewers using standard protocols (22).

Additional questions assessed respondents’ social-occupational class, health status, marital status, smoking, parental cause of death, and alcohol consumption.

Variable handling and statistical analysis

The study outcomes were all-cause mortality and CVD incidence (CVD death or hospital episode). Exposure variables were the total and type-specific participation in physical activity: 1) IDPA was the key exposure variable of interest, comprising heavy housework and heavy gardening/do-it-yourself activities combined; 2) walking for any purpose; and 3) sports and exercise. Physical activity variables were frequency based and were expressed as the number of weekly episodes lasting for at least 20 continuous minutes. Physical activity types were converted into 3 categories: The bottom category was no participation (0 episodes), while the remaining 2 categories were defined by the sex-specific median of weekly number of activity episodes (equal to or below the median (middle category) and over the median (highest category)). When we computed the total physical activity frequency score, there were very few participants with 0 episodes, and therefore the total physical activity frequency score was converted into sex-specific tertiles of weekly activity episodes. Covariables that were entered into the multivariable models were other physical activity types, age, body mass index (weight (kg)/height (m)2) category (≤18.5, >18.5–25, >25–30, >30–35, >35–40, and >40), social-occupational class using the Registrar General classification (I/II, III nonmanual, III manual, IV/V), longstanding illness (yes/no), marital status (single/never married, married, separated/divorced, and windowed), smoking status (never smoked, former smoker, current smoker), parental cause of death (cardiovascular, diabetes, other), alcohol consumption (number of units consumed in the last 7 days), and survey year.

All analyses were carried out in the autumn of 2008 and were sex specific. For individuals who survived and remained CVD free, data were censored to December 2007. The Cox proportional hazards model was used with months as the timescale to estimate the risk of death from any cause or the risk of CVD incident (CVD hospital episode or CVD death) by levels of each physical activity type or total physical activity. The proportional hazards assumption was examined by comparing the cumulative hazard plots grouped on exposure, although no appreciable violations were noted. Test for linear trend was obtained by entering the categorical variables as continuous parameters in the models. We tested a linear trend because our a priori hypothesis was that, if associations between physical activity and our outcomes existed, there would be a dose response of greater benefit with a greater amount of activity. However, where there appeared to be associations that did not follow this trend (e.g., a similar magnitude of association in the middle and highest categories), we compared a model with 2 indicator variables (a model that does not assume a linear trend) with one having the 3 categories entered as a score, using a likelihood ratio test. We used chi-squares to examine the univariable relations of the confounders with the exposure and outcome variables. We tabulated results for age-adjusted, semi-adjusted (age, social-occupational class, marital status), and fully adjusted models (all covariables including mutual adjustments for other activity types). To minimize the chances of reduced physical activity due to prediagnosed/prodromal illness, we excluded respondents who died (all-cause mortality analyses) or had a CVD event (CVD analyses) during the first year of follow-up (as well as excluding from all analyses those with established CVD at baseline). For the same reason, we also excluded respondents with a cancer registration prior to baseline from the all-cause mortality analyses.

Analyses were performed by using SPSS, version 14, software (SPSS, Inc., Chicago, Illinois), and all tests of statistical significance were based on 2-sided probability.

RESULTS

In the core sample, 42.9% of men and 39.8% of women reported participation in IDPA for over 20 minutes per session. Among participators, the mean weekly number of IDPA sessions was 2.3 (standard deviation, 2.2) for men and 2.4 (standard deviation, 2.2) for women. IDPA accounted for 20.1% and 21.2% of men's and women's total physical activity sessions, respectively.

Table 1 shows the associations of total physical activity (all types including IDPA) with covariables. Participants in the highest tertile of total physical activity were more likely to be young and from the manual social-occupational class; less likely to have a parent who died of CVD, to have never been smokers, to have a lower body mass index, and to have a lower all-cause mortality and CVD incidence; and to be more likely to be married (women only), less likely to have longstanding illness, and more likely to report good or very good health.

Table 1.

Descriptive Characteristics of the Cohort at Baseline by Tertiles of Total Weekly Physical Activity Sessions, 1995, 1998, and 2003 Scottish Health Survey Samples

Variable Men
 
Women
 
Total Physical Activitya Tertiles
 
P Value Total Physical Activitya Tertiles
 
P Value 
Bottomb (n = 1,769)
 
Mediumc (n = 2,118)
 
Topd (n = 2,215)
 
Bottomb (n = 3,295)
 
Mediume (n = 1,697)
 
Topf (n = 2,632)
 
Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) 
Activity sessions, no. 3.0 (0.2)  4.5 (0.5)  8.6 (3.8)  <0.001 3.4 (0.5)  5.0 (0.2)  8.0 (2.7)  <0.001 
Age, years 54.3 (11.7)  52.4 (11.5)  48.9 (10.7)  <0.001 55.1 (12.2)  52.4 (11.4)  48.9 (10.4)  <0.001 
Marital status, married  67.3  68.6  67.8 0.68  57.5  60.8  61.4 0.005 
Social class, manual  65.6  60.6  50.8 <0.001  54.8  49.5  43.8 <0.001 
BMI, kg/m2 27.7 (4.8)  27.4 (4.2)  27.12 (3.9)  <0.001 28.0 (5.9)  26.9 (5.2)  26.7 (5.0)  <0.001 
Obesity, BMI ≥30 kg/m2  27.4  24.6  20.0 <0.001  30.5  23.5  20.2 <0.001 
Smoking, never  31.3  36.9  44.3 <0.001  41.0  45.6  47.3 <0.001 
Alcohol intake, units/week 18.1 (26.3)  19.0 (25.2)  16.9 (19.1)  0.029 5.2 (9.4)  5.7 (9.1)  6.7 (8.7)  <0.001 
Family history of CVD, present  30.3  31.7  28.6 0.083  33.8  30.6  28.6 <0.001 
Longstanding illness, present  54.4  44.9  35.2 <0.001  55.5  41.2  37.2 <0.001 
Self-rated health, good  60.1  73.6  84.0 <0.001  63.3  77.1  83.7 <0.001 
Variable Men
 
Women
 
Total Physical Activitya Tertiles
 
P Value Total Physical Activitya Tertiles
 
P Value 
Bottomb (n = 1,769)
 
Mediumc (n = 2,118)
 
Topd (n = 2,215)
 
Bottomb (n = 3,295)
 
Mediume (n = 1,697)
 
Topf (n = 2,632)
 
Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) 
Activity sessions, no. 3.0 (0.2)  4.5 (0.5)  8.6 (3.8)  <0.001 3.4 (0.5)  5.0 (0.2)  8.0 (2.7)  <0.001 
Age, years 54.3 (11.7)  52.4 (11.5)  48.9 (10.7)  <0.001 55.1 (12.2)  52.4 (11.4)  48.9 (10.4)  <0.001 
Marital status, married  67.3  68.6  67.8 0.68  57.5  60.8  61.4 0.005 
Social class, manual  65.6  60.6  50.8 <0.001  54.8  49.5  43.8 <0.001 
BMI, kg/m2 27.7 (4.8)  27.4 (4.2)  27.12 (3.9)  <0.001 28.0 (5.9)  26.9 (5.2)  26.7 (5.0)  <0.001 
Obesity, BMI ≥30 kg/m2  27.4  24.6  20.0 <0.001  30.5  23.5  20.2 <0.001 
Smoking, never  31.3  36.9  44.3 <0.001  41.0  45.6  47.3 <0.001 
Alcohol intake, units/week 18.1 (26.3)  19.0 (25.2)  16.9 (19.1)  0.029 5.2 (9.4)  5.7 (9.1)  6.7 (8.7)  <0.001 
Family history of CVD, present  30.3  31.7  28.6 0.083  33.8  30.6  28.6 <0.001 
Longstanding illness, present  54.4  44.9  35.2 <0.001  55.5  41.2  37.2 <0.001 
Self-rated health, good  60.1  73.6  84.0 <0.001  63.3  77.1  83.7 <0.001 

Abbreviations: BMI, body mass index; CVD, cardiovascular disease; SD, standard deviation.

a

Total physical activity: heavy domestic, walking, and sport sessions.

b

Four or fewer sessions/week.

c

From greater than 4 to 5 sessions/week.

d

Greater than 5 sessions/week.

e

From greater than 4 to 5.5 sessions/week.

f

Greater than 5.5 sessions/week.

Intense domestic physical activity and all-cause mortality

Tables 2 and 3 show the relations between each physical activity type and all-cause mortality risk. Although there was no evidence of a dose-response relation, both men and women who reported doing IDPA (whether above or below the median for women, only below the median for men) had reduced risk of death during the follow-up period compared with those who reported doing no IDPA.

Table 2.

Cox Regression Models for Physical Activity Levels and All-Cause Mortality Among Men Aged 35 or More Years, 1995, 1998, and 2003 Scottish Health Survey Samplesa

 No. of Cases (N = 499) Total (N = 5,896) Age Adjusted
 
Model 1b
 
Model 2c
 
 Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval 
Domestic activity sessions (median)d         
    None (0) 363 3,362 1.00  1.00  1.00  
    Mediume (1.0) 53 1,219 0.47 0.35, 0.62 0.53 0.40, 0.71 0.68 0.50, 0.91 
    Highf (2.5) 83 1,315 0.64 0.50, 0.81 0.71 0.56, 0.91 0.89 0.69, 1.14 
        Ptrend   <0.001 <0.001 0.034 
Walking activity sessions (median)d         
    None (0) 294 3,157 1.00  1.00  1.00  
    Mediume (2.0) 87 1,206 0.87 0.68, 1.10 0.91 0.72, 1.16 1.10 0.86, 1.14 
    Highf (7.0) 118 1,533 0.79 0.64, 0.98 0.77 0.62, 0.95 0.91 0.73, 1.13 
        Ptrend   0.088 0.058 0.394 
Sports activity sessions (median)d         
    None (0) 458 4,502 1.00  1.00  1.00  
    Mediume (1.0) 19 689 0.43 0.27, 0.69 0.51 0.32, 0.80 0.69 0.42, 1.08 
    Highf (2.8) 22 705 0.50 0.32, 0.76 0.53 0.34, 0.82 0.73 0.47, 1.13 
        Ptrend   <0.001 <0.001 0.117 
Total activity sessions (median)dg         
    Low (3.0) 203 1,692 1.00  1.00  1.00  
    Medium (4.8) 178 2,047 0.76 0.62, 0.93 0.80 0.65, 0.97 0.97 0.79, 1.19 
    High (7.0) 118 2,157 0.60 0.47, 0.75 0.65 0.52, 0.82 0.92 0.72, 1.17 
        Ptrend   <0.001 0.001 0.800 
Total activity sessions excluding domestic (median)dg         
    Low (1.0) 258 2,230 1.00  1.00  1.00  
    Medium (3.0) 165 1,567 0.90 0.74, 1.09 0.88 0.72, 1.07 1.03 0.85, 1.26 
    High (4.3) 76 2,099 0.49 0.38, 0.63 0.54 0.42, 0.70 0.76 0.58, 0.98 
        Ptrend   <0.001 <0.001 0.078 
 No. of Cases (N = 499) Total (N = 5,896) Age Adjusted
 
Model 1b
 
Model 2c
 
 Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval 
Domestic activity sessions (median)d         
    None (0) 363 3,362 1.00  1.00  1.00  
    Mediume (1.0) 53 1,219 0.47 0.35, 0.62 0.53 0.40, 0.71 0.68 0.50, 0.91 
    Highf (2.5) 83 1,315 0.64 0.50, 0.81 0.71 0.56, 0.91 0.89 0.69, 1.14 
        Ptrend   <0.001 <0.001 0.034 
Walking activity sessions (median)d         
    None (0) 294 3,157 1.00  1.00  1.00  
    Mediume (2.0) 87 1,206 0.87 0.68, 1.10 0.91 0.72, 1.16 1.10 0.86, 1.14 
    Highf (7.0) 118 1,533 0.79 0.64, 0.98 0.77 0.62, 0.95 0.91 0.73, 1.13 
        Ptrend   0.088 0.058 0.394 
Sports activity sessions (median)d         
    None (0) 458 4,502 1.00  1.00  1.00  
    Mediume (1.0) 19 689 0.43 0.27, 0.69 0.51 0.32, 0.80 0.69 0.42, 1.08 
    Highf (2.8) 22 705 0.50 0.32, 0.76 0.53 0.34, 0.82 0.73 0.47, 1.13 
        Ptrend   <0.001 <0.001 0.117 
Total activity sessions (median)dg         
    Low (3.0) 203 1,692 1.00  1.00  1.00  
    Medium (4.8) 178 2,047 0.76 0.62, 0.93 0.80 0.65, 0.97 0.97 0.79, 1.19 
    High (7.0) 118 2,157 0.60 0.47, 0.75 0.65 0.52, 0.82 0.92 0.72, 1.17 
        Ptrend   <0.001 0.001 0.800 
Total activity sessions excluding domestic (median)dg         
    Low (1.0) 258 2,230 1.00  1.00  1.00  
    Medium (3.0) 165 1,567 0.90 0.74, 1.09 0.88 0.72, 1.07 1.03 0.85, 1.26 
    High (4.3) 76 2,099 0.49 0.38, 0.63 0.54 0.42, 0.70 0.76 0.58, 0.98 
        Ptrend   <0.001 <0.001 0.078 
a

Excludes deaths during the first year of follow-up and cancer registrations and cardiovascular disease events prior to the baseline measurement.

b

Model 1 includes additional adjustment for social economic group and marital status.

c

Model 2 incorporates model 1 with additional adjustment for body mass index category, longstanding illness, self-rated health, family disease history, smoking, and other types of physical activities.

d

The values in parentheses correspond to the median number of sessions per category.

e

The medium category is defined as below the sex-specific median.

f

The high category is defined as above the sex-specific median.

g

The categories for total physical activity correspond to sex-specific tertiles.

Table 3.

Cox Regression Models for Physical Activity Levels and All-Cause Mortality Among Women Aged 35 or More Years, 1995, 1998, and 2003 Scottish Health Survey Samplesa

 No. of Cases (N =  445) Total (N = 7,325) Age Adjusted
 
Model 1b
 
Model 2c
 
 Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval 
Domestic activity sessions (median)d         
    None (0) 337 4,380 1.00  1.00  1.00  
    Mediume (1.0) 47 1,331 0.54 0.40, 0.74 0.56 0.41, 0.77 0.70 0.51, 0.95 
    Highf (3.0) 61 1,614 0.56 0.42, 0.74 0.58 0.42, 0.74 0.70 0.52, 0.93 
        Ptrend   <0.001 <0.001 0.009 
Walking activity sessions (median)d         
    None (0) 266 3,675 1.00  1.00  1.00  
    Mediume (2.0) 88 1,751 0.69 0.54, 0.88 0.72 0.57, 0.91 0.89 0.69, 1.14 
    Highf (7.0) 91 1,899 0.70 0.56, 0.90 0.71 0.51, 0.82 0.92 0.72, 1.18 
        Ptrend   0.001 0.002 0.597 
Sports activity sessions (median)d         
    None (0) 407 5,504 1.00  1.00  1.00  
    Mediume (1.0) 22 901 0.43 0.28, 0.65 0.46 0.30, 0.71 0.59 0.38, 0.91 
    Highf (2.5) 16 920 0.37 0.20, 0.57 0.40 0.24, 0.66 0.57 0.34, 0.95 
        Ptrend   <0.001 <0.001 0.008 
Total activity sessions (median)dg         
    Low (3.0) 281 3,144 1.00  1.00  1.00  
    Medium (5.0) 83 1,628 0.67 0.52, 0.85 0.68 0.53, 0.87 0.84 0.66, 1.08 
    High (7.0) 81 2,553 0.49 0.38, 0.64 0.51 0.40, 0.66 0.71 0.55, 0.93 
        Ptrend   <0.001 <0.001 0.033 
Total activity sessions excluding domestic (median)dg         
    Low (1.0) 254 2,696 1.00  1.00  1.00  
    Medium (3.0) 131 2,065 0.68 0.55, 0.84 0.70 0.56, 0.86 0.87 0.70, 1.08 
    High (4.0) 60 2,564 0.37 0.28, 0.49 0.39 0.29, 0.52 0.57 0.43, 0.77 
        Ptrend   <0.001 <0.001 0.001 
 No. of Cases (N =  445) Total (N = 7,325) Age Adjusted
 
Model 1b
 
Model 2c
 
 Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval 
Domestic activity sessions (median)d         
    None (0) 337 4,380 1.00  1.00  1.00  
    Mediume (1.0) 47 1,331 0.54 0.40, 0.74 0.56 0.41, 0.77 0.70 0.51, 0.95 
    Highf (3.0) 61 1,614 0.56 0.42, 0.74 0.58 0.42, 0.74 0.70 0.52, 0.93 
        Ptrend   <0.001 <0.001 0.009 
Walking activity sessions (median)d         
    None (0) 266 3,675 1.00  1.00  1.00  
    Mediume (2.0) 88 1,751 0.69 0.54, 0.88 0.72 0.57, 0.91 0.89 0.69, 1.14 
    Highf (7.0) 91 1,899 0.70 0.56, 0.90 0.71 0.51, 0.82 0.92 0.72, 1.18 
        Ptrend   0.001 0.002 0.597 
Sports activity sessions (median)d         
    None (0) 407 5,504 1.00  1.00  1.00  
    Mediume (1.0) 22 901 0.43 0.28, 0.65 0.46 0.30, 0.71 0.59 0.38, 0.91 
    Highf (2.5) 16 920 0.37 0.20, 0.57 0.40 0.24, 0.66 0.57 0.34, 0.95 
        Ptrend   <0.001 <0.001 0.008 
Total activity sessions (median)dg         
    Low (3.0) 281 3,144 1.00  1.00  1.00  
    Medium (5.0) 83 1,628 0.67 0.52, 0.85 0.68 0.53, 0.87 0.84 0.66, 1.08 
    High (7.0) 81 2,553 0.49 0.38, 0.64 0.51 0.40, 0.66 0.71 0.55, 0.93 
        Ptrend   <0.001 <0.001 0.033 
Total activity sessions excluding domestic (median)dg         
    Low (1.0) 254 2,696 1.00  1.00  1.00  
    Medium (3.0) 131 2,065 0.68 0.55, 0.84 0.70 0.56, 0.86 0.87 0.70, 1.08 
    High (4.0) 60 2,564 0.37 0.28, 0.49 0.39 0.29, 0.52 0.57 0.43, 0.77 
        Ptrend   <0.001 <0.001 0.001 
a

Excludes deaths during the first year of follow-up and cancer registrations and cardiovascular disease events prior to the baseline measurement.

b

Model 1 includes additional adjustment for social economic group and marital status.

c

Model 2 incorporates model 1 with additional adjustment for body mass index category, longstanding illness, self-rated health, family disease history, smoking, and other types of physical activities.

d

The values in parentheses correspond to the median number of sessions per category.

e

The medium category is defined as below the sex-specific median.

f

The high category is defined as above the sex-specific median.

g

The categories for total physical activity correspond to sex-specific tertiles.

The association of any walking with all-cause mortality was weak in both sexes. Among men, sports and total physical activity were inversely associated with risk for all-cause death, but adjustments for covariables attenuated these associations to the null. This was not the case among women, where the associations of sporting activity and total physical activity (including/excluding IDPA) were attenuated with adjustment for potential confounders, but important associations remained. All physical activity types and total activity remained inversely associated with all-cause mortality in women. In women's models, the magnitude of the point estimates for the top tertile of total physical activity excluding IDPA was stronger than that for the respective point estimate of the score including IDPA.

Intense domestic physical activity and cardiovascular disease events

Tables 4 and 5 show the relations between each physical activity type and CVD risk. The modest inverse association of IDPA with CVD in men attenuated to the null with adjustment for covariables. The inverse association of walking with CVD in women attenuated to the null with adjustment for covariables, although no associations were evident in men. After multivariate adjustments, sports had some limited protective effect in men and less effect in women, where the association was considerably attenuated. Total physical activity (including IDPA) was modestly inversely associated with CVD, but this attenuated to the null with adjustment for covariables. Excluding IDPA from the total number of physical activity sessions resulted in a stronger association with CVD that remained after adjustment for covariables for both men and women (Tables 4 and 5).

Table 4.

Cox Regression Models for Physical Activity Levels and Fatal/Nonfatal Cardiovascular Events Among Men Aged 35 or More Years, 1995, 1998, and 2003 Scottish Health Survey Samplesa

 No. of Cases (N = 468) Total (N = 6,033) Age Adjusted
 
Model 1b
 
Model 2c
 
 Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval 
Domestic activity sessions (median)d         
    None (0) 307 3,432 1.00  1.00  1.00  
    Mediume (1.0) 65 1,252 0.64 0.49, 0.83 0.68 0.52, 0.89 0.82 0.62, 1.08 
    Highf (2.5) 96 1,349 0.82 0.65, 1.04 0.86 0.68, 1.07 0.98 0.77, 1.25 
        Ptrend   0.003 0.016 0.373 
Walking activity sessions (median)d         
    None (0) 262 3,233 1.00  1.00  1.00  
    Mediume (2.0) 76 1,229 0.81 0.63, 1.05 0.84 0.65, 1.08 0.97 0.75, 1.26 
    Highf (7.0) 130 1,571 1.99 0.80, 1.22 0.98 0.80, 1.21 1.10 0.89, 1.37 
        Ptrend   0.276 0.401 0.601 
Sports activity sessions (median)d         
    None (0) 421 4,612 1.00  1.00  1.00  
    Mediume (1.0) 24 704 0.53 0.35, 0.80 0.56 0.37, 0.85 0.70 0.60, 1.06 
    Highf (2.8) 23 717 0.49 0.32, 0.75 0.51 0.34, 0.79 0.67 0.44, 1.03 
        Ptrend   <0.001 <0.001 0.061 
Total activity sessions (median)dg         
    Low (3.0) 169 1,734 1.00  1.00  1.00  
    Medium (4.8) 172 2,096 0.81 0.66, 1.01 0.83 0.67, 1.03 1.01 0.81, 1.26 
    High (7.0) 127 2,203 0.64 0.51, 0.81 0.69 0.54, 0.87 0.90 0.71, 1.15 
        Ptrend   0.001 0.008 0.580 
Total activity sessions excluding domestic (median)dg         
    Low (1.0) 220 2,286 1.00  1.00  1.00  
    Medium (3.0) 162 1,602 0.99 0.81, 1.21 0.98 0.80, 1.21 1.10 0.99, 1.35 
    High (4.3) 86 2,145 0.57 0.44, 0.74 0.61 0.48, 0.79 0.76 0.58, 0.98 
        Ptrend   <0.001 <0.001 0.022 
 No. of Cases (N = 468) Total (N = 6,033) Age Adjusted
 
Model 1b
 
Model 2c
 
 Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval 
Domestic activity sessions (median)d         
    None (0) 307 3,432 1.00  1.00  1.00  
    Mediume (1.0) 65 1,252 0.64 0.49, 0.83 0.68 0.52, 0.89 0.82 0.62, 1.08 
    Highf (2.5) 96 1,349 0.82 0.65, 1.04 0.86 0.68, 1.07 0.98 0.77, 1.25 
        Ptrend   0.003 0.016 0.373 
Walking activity sessions (median)d         
    None (0) 262 3,233 1.00  1.00  1.00  
    Mediume (2.0) 76 1,229 0.81 0.63, 1.05 0.84 0.65, 1.08 0.97 0.75, 1.26 
    Highf (7.0) 130 1,571 1.99 0.80, 1.22 0.98 0.80, 1.21 1.10 0.89, 1.37 
        Ptrend   0.276 0.401 0.601 
Sports activity sessions (median)d         
    None (0) 421 4,612 1.00  1.00  1.00  
    Mediume (1.0) 24 704 0.53 0.35, 0.80 0.56 0.37, 0.85 0.70 0.60, 1.06 
    Highf (2.8) 23 717 0.49 0.32, 0.75 0.51 0.34, 0.79 0.67 0.44, 1.03 
        Ptrend   <0.001 <0.001 0.061 
Total activity sessions (median)dg         
    Low (3.0) 169 1,734 1.00  1.00  1.00  
    Medium (4.8) 172 2,096 0.81 0.66, 1.01 0.83 0.67, 1.03 1.01 0.81, 1.26 
    High (7.0) 127 2,203 0.64 0.51, 0.81 0.69 0.54, 0.87 0.90 0.71, 1.15 
        Ptrend   0.001 0.008 0.580 
Total activity sessions excluding domestic (median)dg         
    Low (1.0) 220 2,286 1.00  1.00  1.00  
    Medium (3.0) 162 1,602 0.99 0.81, 1.21 0.98 0.80, 1.21 1.10 0.99, 1.35 
    High (4.3) 86 2,145 0.57 0.44, 0.74 0.61 0.48, 0.79 0.76 0.58, 0.98 
        Ptrend   <0.001 <0.001 0.022 
a

Excludes existing cardiovascular disease cases at baseline and events during the first year of follow-up.

b

Model 1 includes additional adjustment for social economic group and marital status.

c

Model 2 incorporates model 1 with additional adjustment for body mass index category, longstanding illness, self-rated health, family disease history, smoking, and other types of physical activities.

d

The values in parentheses correspond to the median number of sessions per category.

e

The medium category is defined as below the sex-specific median.

f

The high category is defined as above the sex-specific median.

g

The categories for total physical activity correspond to sex-specific tertiles.

Table 5.

Cox Regression Models for Physical Activity Levels and Fatal/Nonfatal Cardiovascular Events Among Women Aged 35 or More Years, 1995, 1998, and 2003 Scottish Health Survey Samplesa

 No. of Cases (N = 326) Total (N = 7,561) Age Adjusted
 
Model 1b
 
Model 2c
 
 Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval 
Domestic activity sessions (median)d         
    None (0) 229 4,536 1.00  1.00  1.00  
    Mediume (1.0) 44 1,370 0.73 0.59, 1.15 0.76 0.55, 1.05 0.92 0.66, 1.28 
    Highf (3.0) 53 1,655 0.70 0.55, 1.04 0.70 0.51, 0.95 0.86 0.63, 1.18 
        Ptrend   <0.001 0.034 0.638 
Walking activity sessions (median)d         
    None (0) 195 3,805 1.00  1.00  1.00  
    Mediume (2.0) 59 1,799 0.63 0.47, 0.85 0.65 0.49, 0.88 0.80 0.59, 1.08 
    Highf (7.0) 72 1,957 0.75 0.57, 0.99 0.75 0.57, 0.99 0.95 0.61, 1.26 
        Ptrend   0.004 0.007 0.337 
Sports activity sessions (median)d         
    None (0) 286 5,688 1.00  1.00  1.00  
    Mediume (1.0) 24 926 0.64 0.45, 1.06 0.67 0.44, 1.01 0.82 0.54, 1.25 
    Highf (2.5) 16 947 0.48 0.30, 0.85 0.52 0.31, 0.86 0.71 0.42, 1.18 
        Ptrend   0.003 0.009 0.301 
Total activity sessions (median)dg         
    Low (3.0) 184 3,254 1.00  1.00  1.00  
    Medium (5.0) 70 1,686 0.82 0.62, 1.08 0.83 0.63, 1.10 1.04 0.78, 1.38 
    High (7.0) 72 2,621 0.58 0.44, 0.76 0.60 0.45, 0.80 0.81 0.61, 1.08 
        Ptrend   0.001 0.002 0.264 
Total activity sessions excluding domestic (median)dg         
    Low (1.0) 177 2,787 1.00  1.00  1.00  
    Medium (3.0) 90 2,127 0.62 0.49, 0.81 0.65 0.50, 0.84 0.80 0.61, 1.03 
    High (4.0) 59 2,647 0.46 0.34, 0.62 0.49 0.36, 0.67 0.68 0.50, 0.93 
        Ptrend   <0.001 <0.001 0.033 
 No. of Cases (N = 326) Total (N = 7,561) Age Adjusted
 
Model 1b
 
Model 2c
 
 Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval Hazard Ratio 95% Confidence Interval 
Domestic activity sessions (median)d         
    None (0) 229 4,536 1.00  1.00  1.00  
    Mediume (1.0) 44 1,370 0.73 0.59, 1.15 0.76 0.55, 1.05 0.92 0.66, 1.28 
    Highf (3.0) 53 1,655 0.70 0.55, 1.04 0.70 0.51, 0.95 0.86 0.63, 1.18 
        Ptrend   <0.001 0.034 0.638 
Walking activity sessions (median)d         
    None (0) 195 3,805 1.00  1.00  1.00  
    Mediume (2.0) 59 1,799 0.63 0.47, 0.85 0.65 0.49, 0.88 0.80 0.59, 1.08 
    Highf (7.0) 72 1,957 0.75 0.57, 0.99 0.75 0.57, 0.99 0.95 0.61, 1.26 
        Ptrend   0.004 0.007 0.337 
Sports activity sessions (median)d         
    None (0) 286 5,688 1.00  1.00  1.00  
    Mediume (1.0) 24 926 0.64 0.45, 1.06 0.67 0.44, 1.01 0.82 0.54, 1.25 
    Highf (2.5) 16 947 0.48 0.30, 0.85 0.52 0.31, 0.86 0.71 0.42, 1.18 
        Ptrend   0.003 0.009 0.301 
Total activity sessions (median)dg         
    Low (3.0) 184 3,254 1.00  1.00  1.00  
    Medium (5.0) 70 1,686 0.82 0.62, 1.08 0.83 0.63, 1.10 1.04 0.78, 1.38 
    High (7.0) 72 2,621 0.58 0.44, 0.76 0.60 0.45, 0.80 0.81 0.61, 1.08 
        Ptrend   0.001 0.002 0.264 
Total activity sessions excluding domestic (median)dg         
    Low (1.0) 177 2,787 1.00  1.00  1.00  
    Medium (3.0) 90 2,127 0.62 0.49, 0.81 0.65 0.50, 0.84 0.80 0.61, 1.03 
    High (4.0) 59 2,647 0.46 0.34, 0.62 0.49 0.36, 0.67 0.68 0.50, 0.93 
        Ptrend   <0.001 <0.001 0.033 
a

Excludes existing cardiovascular disease cases at baseline and events during the first year of follow-up.

b

Model 1 includes additional adjustment for social economic group and marital status.

c

Model 2 incorporates model 1 with additional adjustment for body mass index category, longstanding illness, self-rated health, family disease history, smoking, and other types of physical activities.

d

The values in parentheses correspond to the median number of sessions per category.

e

The medium category is defined as below the sex-specific median.

f

The high category is defined as above the sex-specific median.

g

The categories for total physical activity correspond to sex-specific tertiles.

DISCUSSION

There is a paucity of data examining the independent health benefits of IDPA. In this study, IDPA appeared to be independently associated with decreased risk for all-cause mortality in a representative sample of adults living in Scotland. An English study of approximately 15,000 adults (23) also found a protective effect of domestic activity on all-cause mortality. This suggests that, regardless of its cardioprotective qualities, IDPA may play an important role in preventing premature death, perhaps through better general health and through the prevention of other conditions, such as cancer. Household activity has been found to confer the strongest of all activity types and protection from breast and endometrial cancer in 2 recent reports (24, 25) that utilized large samples (n = >215,000) of European women.

We found that IDPA was largely unrelated to the risk for CVD events and that including IDPA weakened the relations between total activity and such events. The few cross-sectional studies (7, 9, 14) that investigated the relation between CVD risk factors and IDPA are in concordance with these results. Two studies that looked at the independent effects of IDPA on CVD mortality (23, 26) contradict our results in that activities such as gardening and repair work (26) or “home activity” (23) showed protective effects against cardiovascular mortality among samples of men from rural Finland and men and women from rural England, respectively. For comparison with both of these studies (23, 26) that examined fatal CVD events only, we carried out an alternative analysis by CVD event type (fatal vs. nonfatal). The results offered some indication of an effect of IDPA on fatal but not on nonfatal CVD events, but we did not have sufficient power to examine this question in full. Further, gardening in rural Finland and IDPA in rural England may be more intense than IDPA in primarily urban (>65%) Scotland. That is, the components of IDPA in rural environments may involve more intense gardening and outdoor do-it-yourself activities, such as digging, building work, chopping wood, and less indoor housework, such as cleaning windows and doing the laundry, which are of lower intensity. It could be that IDPA needs to be more closely defined; for example, intense gardening may be related to a reduced risk for CVD, but intense housework may not. In an alternative analysis that we carried out, neither of the individual components of IDPA was related to CVD events. Our definition of IDPA was very different from “home activity” in the English study (23). For example, this study included stair climbing, an activity with cardioprotective qualities (11). Stair climbing was not part of IDPA in our study. The finding that, once the authors excluded stair climbing and digging, the associations between home activity and mortality disappeared (23) adds plausibility to the above explanations.

IDPA is cited in public health recommendations (3–5) because of estimates that the energy cost is equivalent to moderate-intensity activity of 3.0–6.0 metabolic equivalents, which is the minimum intensity required to achieve a health benefit (20). Most types of IDPA are classified as moderate intensity (9) and are expected to have health benefits equal to other moderate-intensity activity, such as walking and cycling. While these latter types are characterized by the use of large muscle groups utilized in a rhythmic and dynamic nature (20), many housework activities (e.g., cleaning windows, vacuum cleaning) are performed in a restricted household space and utilize smaller upper body muscles (27) or are intermittent, less rhythmic, and often nonlocomotor, and this may partly explain the lack of an apparent protective effect on CVD. Another possible explanation is that the Scottish Health Survey questions capture light-intensity domestic activity as well as heavy intensity, diluting the true relations between exposure and outcome.

If frequency alone is an inadequate indicator and total volume (the sum of the frequency and duration) of activity is what confers protection against future disease, our reporting of physical activity in terms of the frequency of ≥20-minute episodes might have led to misclassification. We reanalyzed the Scottish Health Survey 1998 and 2003 samples, where detailed duration information was available. The mean duration of intense domestic physical activity sessions was longer (71 (standard deviation, 83) minutes) than that of walking (52 (standard deviation, 48) minutes) (P < 0.001). Sports session durations varied by sport type but in general were considerably shorter than IDPA. Consequently, although our analysis utilized frequency but not duration of activity sessions, it is unlikely that this introduced a bias. Despite these measurement limitations, it remains a possibility that IDPAs are not of a sufficient intensity to be cardioprotective and may not represent a moderate intensity. We restricted the analysis to those aged over 60 years, for whom IDPA might represent a higher relative intensity but results were also nonsignificant.

In a recent meta-analysis (28), walking was associated with a 31% reduction in CVD risk, although we did not find such a strong relation in the present analysis. A possible explanation for this is that we included walking of any intensity in our analysis because of incompatibility of the intensity questions between the 1995 and 1998/2003 questionnaires. In the Scottish Health Survey 1998 and 2003 samples where there was a question on usual walking pace, approximately 70% of walking was reported to be of slow or average pace that most likely corresponded to submoderate intensities. Given that a brisk walking pace is associated with greater risk reduction, it is likely that slow/moderate paced walking is not of sufficient intensity to elicit benefit, and this may be responsible for the aforementioned lack of relations.

The strengths of this study include the large, representative, and heterogeneous sample of the Scottish population and the availability of many potential confounders. The availability of a “hard” outcome such as clinically confirmed hospital episodes and the subsequent exclusion of existing CVD and cancer cases minimized the risk of reverse causality.

Our study has certain limitations. Self-reported physical activity measures have inherently large measurement error (29). If there was reporting bias and participants selectively overreported IDPA, then the true effect of IDPA on CVD risk would be underestimated. We included only nonoccupational physical activity. Although we adjusted for occupational-derived social class, it remains a possibility that occupational activity levels may have influenced our results. Because the Scottish population has a very high rate of coronary heart disease (1), we also acknowledge that there may be unmeasured confounders, such as diet. We would anticipate that diet has a similar confounding effect with each activity type and so would not affect our main aim. Physical activity was assessed only at baseline, and this may have affected the precision of our estimates because physical activity levels decrease with age (30). There are indications that such decreases may dilute the relations between mortality and physical activity up to 60% (31).

In conclusion, IDPA seems to offer some protection against all-cause mortality, but its cardioprotective benefits are questionable. IDPA participation for CVD prevention is widely promoted, and this may displace other more cardiovascular health-enhancing physical activity because of time constraints. Although our study supports the promotion of IDPA for general health benefits, it might be sensible for practitioners and clinicians specializing in CVD prevention to encourage individuals to adhere to the current physical activity recommendations, mainly by means of continuous and rhythmic forms (20) of moderate-to-vigorous physical activity. Further large prospective studies are required to establish the specific health benefits of each type of physical activity so that primary care, clinical, and public health physical activity recommendations are based on solid evidence.

Abbreviations

    Abbreviations
  • CVD

    cardiovascular disease

  • IDPA

    intense domestic physical activity

Author affiliations: Department of Epidemiology and Public Health, University College London, London, United Kingdom (Emmanuel Stamatakis, Mark Hamer); Department of Social Medicine, University of Bristol, Bristol, United Kingdom (Debbie A. Lawlor); and Medical Research Council Centre for Causal Analyses in Translational Epidemiology, University of Bristol, Bristol, United Kingdom (Debbie A. Lawlor).

The Scottish Health Survey is funded by the Scottish Executive.

Conflict of interest: none declared.

References

1.
British Heart Foundation
Tables 1.6 and 2.2
Coronary Heart Disease Statistics 2007
 , 
2007
London, United Kingdom
British Heart Foundation
2.
Wartburton
DE
Nicol
CW
Bredin
SD
Health benefits of physical activity: the evidence
Can Med Assoc J
 , 
2006
, vol. 
174
 
6
(pg. 
801
-
809
)
3.
The Chief Medical Officer
At Least Five a Week: Evidence on the Impact of Physical Activity and Its Relationship to Health
 , 
2004
London, United Kingdom
Department of Health
4.
President's Council on Physical Fitness and Sports, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, US Department of Health and Human Services
Physical Activity and Health: A Report of the Surgeon General
 , 
1996
Atlanta, GA
Centers for Disease Control and Prevention
5.
Haskell
WL
Lee
IM
Pate
RR
, et al.  . 
Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association
Med Sci Sports Exerc.
 , 
2007
, vol. 
39
 
8
(pg. 
1423
-
1434
)
6.
Yu
S
Yarnell
JW
Sweetnam
PM
, et al.  . 
What level of physical activity protects against premature cardiovascular death? The Caerphilly study
Heart
 , 
2003
, vol. 
89
 
5
(pg. 
502
-
506
)
7.
Andersen
LB
Schnohr
P
Schroll
M
, et al.  . 
All-cause mortality associated with physical activity during leisure time, work, sports, and cycling to work
Arch Intern Med.
 , 
2000
, vol. 
160
 
11
(pg. 
1621
-
1628
)
8.
Fransson
E
de Faire
U
Ahlbom
A
, et al.  . 
The risk of acute myocardial infarction—interactions of types of physical activity
Epidemiology
 , 
2004
, vol. 
15
 
5
(pg. 
573
-
582
)
9.
Stamatakis
E
Hillsdon
M
Primatesta
P
Domestic physical activity in relationship to multiple CVD risk factors
Am J Prev Med.
 , 
2007
, vol. 
32
 
4
(pg. 
320
-
327
)
10.
Fransson
EI
Alfredsson
LS
de Faire
UH
, et al.  . 
Leisure time, occupational and household physical activity, and risk factors for cardiovascular disease in working men and women: the WOLF Study
Scand J Public Health
 , 
2003
, vol. 
31
 
5
(pg. 
324
-
333
)
11.
Boreham
CA
Kennedy
RA
Murphy
MH
, et al.  . 
Training effects of short bouts of stair climbing on cardiorespiratory fitness, blood lipids, and homocysteine in sedentary young women
Br J Sports Med.
 , 
2005
, vol. 
39
 
9
(pg. 
590
-
593
)
12.
Barengo
NC
Hu
G
Lakka
TA
, et al.  . 
Low physical activity as a predictor for total and cardiovascular disease mortality in middle aged men and women in Finland
Eur Heart J.
 , 
2004
, vol. 
25
 
24
(pg. 
2204
-
2211
)
13.
Bucksch
J
Helmert
U
Leisure time sports activity and all-cause mortality in West Germany (1984–1998)
J Public Health
 , 
2004
, vol. 
12
 (pg. 
351
-
358
)
14.
Lawlor
DA
Taylor
M
Bedford
C
, et al.  . 
Is housework good for health? Levels of physical activity and factors associated with activity in elderly women. Results from the British Women's Heart and Health Study
J Epidemiol Community Health
 , 
2002
, vol. 
56
 
6
(pg. 
473
-
478
)
15.
Weller
I
Corey
P
The impact of excluding non-leisure energy expenditure on the relation between physical activity and mortality in women
Epidemiology
 , 
1998
, vol. 
9
 
6
(pg. 
632
-
635
)
16.
Bucksch
J
Physical activity of moderate intensity in leisure time and the risk of all cause mortality
Br J Sports Med.
 , 
2005
, vol. 
39
 
9
(pg. 
632
-
638
)
17.
Matthews
CE
Jurj
AL
Shu
XO
, et al.  . 
Influence of exercise, walking, cycling, and overall nonexercise physical activity on mortality in Chinese women
Am J Epidemiol.
 , 
2007
, vol. 
165
 
12
(pg. 
1343
-
1350
)
18.
Health Education Authority
Promoting Physical Activity in Primary Care
 , 
1996
London, United Kingdom
Health Education Authority
19.
National Health Service
Our Healthier Nation, National Service Framework for Coronary Heart Disease
 
London
United Kingdom
 
The Stationery Office; 2000
20.
American College of Sports Medicine
Position Stand: the recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness, and flexibility in healthy adults
Med Sci Sports Exerc.
 , 
1998
, vol. 
30
 
6
(pg. 
975
-
991
)
21.
The Scottish Executive
The Scottish Health Survey—2003 Results
 , 
2005
Edinburgh, United Kingdom
The Scottish Executive
 
22.
Joint Health Surveys Unit
Health Survey for England Physical Activity Validation Study: Substantive Report
 , 
2007
Leeds, United Kingdom
Information Centre for Health and Social Care
23.
Besson
H
Ekelund
U
Brage
S
, et al.  . 
Relationship between subdomains of total physical activity and mortality
Med Sci Sports Exerc.
 , 
2008
, vol. 
40
 (pg. 
1909
-
1915
)
24.
Lahmann
PA
Friedenreich
C
Schit
AJ
, et al.  . 
Physical activity and breast cancer risk: the European Investigation into Cancer and Nutrition
Cancer Epidemiol Biomarkers Prev.
 , 
2007
, vol. 
16
 
1
(pg. 
36
-
42
)
25.
Friedenreich
C
Cust
A
Lahmann
PA
, et al.  . 
Physical activity and endometrial cancer: the European Investigation into Cancer and Nutrition
Int J Cancer.
 , 
2007
, vol. 
121
 
2
(pg. 
347
-
355
)
26.
Haapanen
N
Miilunpalo
S
Vuori
I
, et al.  . 
Characteristics of leisure time physical activity associated with decreased risk of premature all-cause and cardiovascular mortality in middle-aged men
Am J Epidemiol.
 , 
1996
, vol. 
143
 
9
(pg. 
870
-
880
)
27.
Rosano
A
Moccaldi
R
Cioppa
M
, et al.  . 
Musculoskeletal disorders and housework in Italy. (In Italian)
Ann Ig
 , 
2004
, vol. 
16
 
3
(pg. 
497
-
507
)
28.
Hamer
M
Chida
Y
Walking and primary prevention: a meta-analysis of prospective cohort studies
Br J Sports Med.
 , 
2008
, vol. 
42
 
4
(pg. 
238
-
243
)
29.
Matthews
CE
Welk
GJ
Use of self-report instruments to assess physical activity
Physical Activity Assessments for Health-related Research
 , 
2002
Champaign, IL
Human Kinetics
30.
Joint Health Surveys Unit
Cardiovascular Disease Joint Risk Factors
 , 
2008
Leeds, United Kingdom
Health and Social Care Information Centre
 
The Health Survey for England 2006; vol 1
31.
Andersen
LB
Relative risk in the physically inactive is underestimated because of real changes in exposure level during follow-up
Am J Epidemiol.
 , 
2004
, vol. 
160
 
2
(pg. 
189
-
195
)